For AI agents: the complete documentation index is available at https://docs.us.ovhcloud.com/en/llms.txt, the full documentation bundle is available at https://docs.us.ovhcloud.com/en/llms-full.txt, and this page is available as Markdown at https://docs.us.ovhcloud.com/en/public-cloud/ai-machine-learning/ai-training-capabilities.md.
  • English

AI Training - Features, capabilities, and limitations

View as Markdown

Learn the technical capabilities and limitations of the Public Cloud AI Training offer.

Info

AI Training is covered by OVHcloud Public Cloud Special Conditions.

 

Features

Available features

AI Training allows you to train your models easily, with just a few clicks or commands. This solution runs your training job on the computational cloud resources you have chosen (GPU/CPU).

FeatureDetails
AI environments
Preinstalled Machine Learning environmentsAI Training comes with a generic Python environment (Conda) or pre-installed ones, such as Pytorch, Tensorflow, FastAI, and HuggingFace.
Easy customizationAI Training allows installation of almost any Conda or Pip packages. You can easily customize your environment to suit your needs. It also allows you to use your own environment by specifying your Docker image.
Management
Multiple ways to manage your training jobsYou can manage your AI Training jobs through the OVHcloud Control Panel, API, or CLI. Depending on your needs, you can easily automate their creation and deletion as well.
Easy start and StopYou can start and stop a training job in one click.
Check the logs of your jobYou can also check the logs of your training job through the OVHcloud Control Panel, API, or CLI.
Compute resources
Guaranteed compute resourcesSelect the number of CPUs or GPUs required during the creation of the AI Training job. Once launched, you will keep these resources as long as your job is running.
Background executionYour tasks can be executed in the background, meaning that closing your web browser will have no effect on your work.
No maximum runtimeYour tasks can last as long as your job is running.
Monitoring toolsEach AI Training service comes with a native Grafana dashboard, allowing you to keep track and monitor your CPU, GPU, RAM and storage resources.
Storage
Fast and flexible storageEach AI Training service comes with local storage, but also the ability to attach remote storage from Object Storage. From a few GiB to multiple TiB, we push your data near our compute power on fast SSD storage for better performances.
Git repositories importationDuring the creation of your AI Training job, you can specify one or multiple Git repositories to download inside your job environment.
Security
Open or restricted authenticationDuring the creation of your AI Training job, select open or restricted access to your job. If restricted, people can be granted access via token or credentials to securely access your environment.
Billing
Easy billingYou only pay for what you consume, billed per minute with per-second granularity.

 

Logs and Monitoring tools

Command line interface (CLI)

AI Training is compliant with the OVHcloud AI CLI. Discover how to install the OVHcloud AI CLI.

Logs

To check the logs of your job, you can do it via the ovhai CLI using the following command:

ovhai job logs <job-id>
Monitoring tools

To see information about your job, you can do so with the ovhai CLI using the command above:

ovhai job get <job-id>

You can then access your metrics through the Monitoring Url.

You are also able to check it from the OVHcloud Control Panel in your job information by clicking the Go to Graph Dashboard button.

 

Planned features

We continuously improve our offers. You can follow, vote and submit ideas to add to our roadmap.

 

Capabilities and limitations

Supported regions for jobs

AI Training can be used from any country in the world, as long as you have an OVHcloud account. Our AI & Machine Learning services are based in the US-EAST-VA (Vint Hill, Virginia) region.

 

Attached resources

Compute resources

You can either choose the number of GPUs or CPUs for an AI Training job, not both. By default, a job uses one GPU. The memory resource is not customizable.

If you choose GPU:

  • CPU, memory and local storage resources are not customizable but scaled linearly with each additional GPU.

If you choose CPU:

  • Memory and local storage resource is not customizable but scaled linearly with each additional CPU.

The maximum amount of CPU/GPU, memory per CPU/GPU and local storage is available on the OVHcloud website, Control Panel and the ovhai CLI.

ovhai capabilities flavor list

For your information, the current limits are:

  • CPU: 12 per job.
  • GPU: 4 per job.
Available hardware for AI Training

Currently, we provide:

  • ai1-1-cpu
  • L4-1-gpu
  • L40s-1-gpu

 

Available storage

Local storage

Each AI Training job comes with a local storage space, which is ephemeral. When you delete your job, this storage space is also deleted. This storage space depends on the selected instances during your AI Training job creation. Please refer to the compute resources section for more information.

Info

Local storage is limited and not the recommended way to handle data, see the OVHcloud documentation on data for more information.

Attached storage

You can attach data volumes from Public Cloud Object Storage. For best performance, we recommend keeping the Object Storage bucket in the same region as your AI Training job. Mounting a bucket from a different region also works, but adds cross-region network latency. Attached storage allows you to work on several TB of data, while being persistent when you delete your AI Training job.

 

Maximum execution time

Warning

NOTE: There is no duration limitation on AI Training job execution.

However, by default, your AI Training job will automatically shut down after 7 consecutive days of being in a RUNNING state. You can also enable Automatic Restart to have it restart every 7 days, or contact our support to extend this period to 28 days.

 

Pre-installed AI environments

OVHcloud AI Training comes with pre-installed AI environments.

List of available AI Environments:

  • AutoGluon + MXNet
  • FastAI
  • HuggingFace
  • MXNet
  • One image to rule them all
  • PyTorch
  • TensorFlow2

 

Environment customization

Each environment can be customized directly with PIP or CONDA (we support almost any package and library).

You can also use your own Docker images.

Docker images can be hosted in a public or private registry.

The use of docker-compose is not possible.

Please be aware that images need to be built with an AMD architecture.

Learn how to build and use your custom Docker image in this tutorial.

 

Network

  • Public networking can be used for all the AI Tools.
  • Private networking (OVHcloud vRack) is not supported.

 

Available ports to public network

Each job has a public URL; by default, this URL accesses the port 8080 of the job. The default port can be configured when you submit a new job.

You can also access other ports by appending them to the URL. For example, the job URL for accessing the port 9000 (starting with the job's ID followed by the port number):

 

Quotas per Public Cloud project

Each Public Cloud project grants a customer, by default, a maximum of 4 GPUs simultaneously. Reach out to our support if you need to increase this limitation.

 

Go further

Browse all the AI & Machine Learning guides.

Browse the rest of the OVHcloud US documentation for guides on our other products and services.

Learn more about AI & Machine Learning on the OVHcloud US website.

If you need training or technical assistance to implement our solutions, contact your sales representative or click on this link to get a quote and ask our Professional Services experts for a custom analysis of your project.